- Anthropic is investing $30 billion in Azure compute capacity for future Claude deployments
- The agreement involves an optional expansion of the infrastructure capacity to a maximum of one gigawatt.
- Future models will train with Nvidia Grace Blackwell and Vera Rubin systems
Anthropic has agreed to acquire $30 billion in computing capacity in Microsoft Azure as part of an expanded strategic partnership between Microsoft and Nvidia.
HE agreement includes the ability to obtain additional capacity of up to one gigawatt for future models based on Nvidia Grace Blackwell and Vera Rubin systems.
The agreement allows Anthropic’s Claude models to scale on Azure-based infrastructure, while enterprise customers can choose Claude within Microsoft’s enterprise platforms.
Access to all Microsoft business products
The collaboration is part of Anthropic’s ongoing plan to operate and train its systems in large clusters optimized for high-performance workloads.
Claude Templates are available to Microsoft Foundry customers and provide access to Claude Sonnet 4.5, Claude Opus 4.1 and Claude Haiku 4.5.
The company said the expansion would make Claude the only distributed border model among major global cloud providers.
Azure customers can access these templates in addition to existing options and leverage Claude-specific features to develop workflows with AI tools and integration into enterprise systems.
Claude also remains available within the Microsoft Copilot suite of products, including GitHub Copilot, Microsoft 365 Copilot, and Copilot Studio.
The deal includes financial commitments from Microsoft and Nvidia, which plan to invest up to $5 billion and $10 billion, respectively, in Anthropic.
The companies confirmed that the collaboration goes beyond funding through a technical partnership aimed at optimizing anthropic workloads on current and future Nvidia architectures.
The goal is to improve performance and efficiency while reducing operating costs in the long term.
Anthropic also wants to ensure that its platforms benefit from hardware developments tailored for large-scale distributed computing.
This agreement allows Anthropic to expand its deployment into enterprise environments that rely heavily on cloud hosting and storage.
The scope of the agreement suggests a strategic initiative to ensure continued access to computing capacity, which could mitigate the constraints associated with rapid model development.
The partnership could also allow Microsoft and Nvidia to strengthen their positions in competitive AI markets through extensive ecosystem adaptation rather than siled product integration.